I have the pleasure of running the last panel today. I would like to ask a quick question. Actually, on behalf of the entire panel who has AI fatigue. Okay. So what we thought we'd do is end the day on AI. To help me do that, I would like to introduce Jon from Pet Circle and Maria from Unilever. Why don't we do a little bit of a round of introductions, maybe your name, where you're from and a little bit about, you know, where you've been in your career, if that's okay. Jon, we'll start with you. So just testing. Okay. I am Jon. I'm the Chief Growth and Marketing Officer at Pet Circle. I've been in digital marketing for longer than I care to admit, but for the last five years I've been working at Pet Circle, we're probably the largest e-com player that you've never heard of, I think we're the third largest by revenue. You know, we service almost a million customers annually. And what we're really trying to do is we're trying to create an experience where we can map the right product to every kind of pet species breed need. And so we're really trying to map the customer to the infinite amount of choice that exists out there. And of course, because we're an online, I can carry a lot more range than perhaps the brick and mortar competitors. And actually, the reason I ended up in digital is because 20 years ago I figured this thing called Google and Internet was going to be have a profound impact on marketing, I think AI is going to have even more of an effect. I think you can take AI, mobile and every other major societal event in the last 30 years, and A.I. is going to have that kind of impact. And maybe leave you with this intro. I think you got to have two mindsets when you're thinking about AI. First is you got to flip the narrative. So this is not the end of creativity. It's the start of infinite, more creative possibilities. Now, this is not the end of my teams. It's actually how you're going to actually enablement, empower them to do more. So I think you've got to really have a mindset that changes. This is going to enable a ton of stuff. And then, you know, I think the you know, the other so the other sort of key thing is you've got to start simple. It's really complex. So get out there And do I mean, I think I mean, one of the points made is you got to keep things simple because it's all about execution, you know, So and then I think actually the third I will add a third one is you got to educate yourself. There's a lot of information, a lot of misinformation. You know, I'm going back to school. I'm trying to learn as much about A.I. as I possibly can because it's complex and it's difficult to know where to start. So go out there and, you know, there's plenty of good, plenty of information and find out as much as you possibly can. Thanks, Jon. Maria, can we hear from you? Hard to come out like I'm after this big introduction, but I'll do my best. I'm Maria. I lead the digital marketing and data team for Unilever and I also lead the AI acceleration team for our ANZ business. I am not from Australia. I'm from Argentina. So I came to this country four years ago with my company with Unilever, and I been working at Unilever for the past five years here and in Latin America, and I do all of my career across marketing, digital marketing, data, a bit of sales and now dipping my toes into the retail media space and also the innovation across AI. I agree with Jon. I feel like we are living in this amazing era to increase our digital literacy, our AI literacy. So I'm really excited on everything that is happening and how we can use AI to do good things, to keep on doing good business and to keep on delivering on sustainable growth for our market. I think the first thing to acknowledge is that AI is not new. Silvana started today by acknowledging actually SAP Emarsys' has had AI in the platform since 2012. What I think a lot of us are really thinking about now is Generative AI, which is really the next iteration. And I'm going to steal a chain from Oroton's term, which I loved about being the author of the next chapter. Right. And Jon, as you just said, removing the ego and really thinking and being very intentional about what it means for your business and actually learning about the capability and the posture is super important. So my first real question is how is your business approaching AI top-down, bottom-up. How are you thinking about it and the approaches that you're putting in place? So I'll to acknowledge the first question I think it's got to be top-down. We tried initially trying to seed AI projects into the teams and through for many reasons. We just could not we could not get traction. I mean, there's the mindset. One, you know, if I adopt AI, I'm going to have a small team. There's a complexity piece. You know, there's so much information and then, you know, a day job. And so what we've done is we've made a top-down, we've got measures. We have a P1 system is one of my P1s. So I've actually got metrics and goals to get to drive AI into my areas. So that's one critical piece. I think the other piece is data. I mean, you've got to get your data right. It's been a lot. That's where you've got to spend a lot of time. In my business, we've got, we think about three tranches of data. And we have this debate. Who's our customer? Is it the person who buys the pet food and the pet stuff or is it the pet? Actually, we think it's the pet. So the most critical piece of the data. Three things I need to know. I really need to know two things plus one. It's nice. Need to know the breed, need to know how old it is. And ideally, I'd like to know the name so that we can start talking them personally. With those three things, I can make great recommendations. I can allow our customers to make great choices. The next piece is product information data. We go through this really laborious task now of re-tagging everything so that we can then eventually we can map these two things together. Right? So if I know you've got a Labrador, we know that I have joint issues. We also know that glucosamine is really good for joint issues. So I can then recommend food that has it. If I've done my tagging and my structures correctly so I can infer... I would actually, before you start getting into Generative AI and all this stuff, get your data right, think about how you're going to use it and get it right. And then the third, the least important piece of data is actually the person buying stuff, you know? But we need to know who they are. We need to know where they are, the demographics and all that stuff. But for me, the map of the pet with the product is really, really critical. Maria huge organization globally. I'm sure there's not a consumer that you haven't touched in terms of the product reach. How are you thinking about AI and the approach and the structure that you're thinking in a business? So in Unilever globally there's so much AI going on, it's probably we have a lot of fatigue for sure. There are over 500 initiatives running on AI across a multiplicity of the disciplines that we touch on. The approach is basically like two fold. We have sort of like a deep approach in innovation. So everything that AI can facilitate to provide a competitive advantage about the bandage. So projects that are like more IP protected, that are changing the way that we operate within our business, and then we have sort of like a more broad implementation that is what are the actual tools that everyone can use and we can sort of make available, train our teams, generate information towards that. But we have a very high ethical mindset on how we use AI and what is the data that we put in place. So of course putting our data together and making everything accessible, organized taxonomies and all of that, it's in place. So that's sort of like evergreen really, and basics. We go very like strict in how we sort of create and name our data sets. And then there is different approach around like specific implementations of tools. So yeah, like basic use of tools for productivity, for marketing, for data implementation. And then there is a massive work on innovation. And I think that in our cases, both like top-down and bottom-up, we have a page where we can sort of pitch for budgets and ideas and like new initiatives that we might want to use. And then there is a massive top-down from our global CEO on like embracing AI for productivity basis. So we are in kind of this like nice moment where everyone is sort of like getting excited and the people that is a bit afraid is getting trained to like sort of prevent that space. I think it's okay to be afraid. Yeah. And it's okay to normalize that. It's getting to another level of practicality than you said, starting to implement some some specific things. So in the lens of because AI can imagine a lot of different things. So worker productivity versus actual marketing. So in the context of today, what are some of the actual use cases and things that you're implementing now and the areas of focus? Yeah. So for me, everything that is happening on the digital commerce space is super important. Like how do you make sure that our omnichannel assets are created? Well, that's something that used to be very expensive and very complex for a company that has thousands of SKUs in market. Now we are using a lot of tools and AI to like gen AI on images to ensure that we have sort of like omnichannel coverage like perfectly done. We have a lot of productivity, of course, I feel like everyone is using Copilot or yeah, we are using Copilot like on our day to day basis to do like things that used to take me hours because I'm not from here and I'm hesitating my language a lot. Now copilot help me to do it better. And then we are working a lot on specifics of our campaign. So we have a very strong commitment to sort of expand the AI literacy into our consumers. So, for example, we have we just did a campaign that had a playbook to better promptings or real beauty prompting that is a tool that we are sharing in general, trying to educate everyone It's in there to avoid being bias and stereotype or preventing bad use. So yeah, lots lots going on from a marketing perspective. Lots going on. Jon same question. What are some of the sort of practical use cases or areas of focus that you're looking at? Yeah, the way that the top-down framework, we sort of split out the application into five areas that's content, customer service, productivity, personalization and analytics. So we've got owners for each of those areas. So to break down what we're doing. In our customer service as an example, keeping things simple. We we're going to just focus on the use case of where's my order when we get it wrong. It really, really pissed customers. Like now we've participated this unconditional bond between pet and pet parent. It's great, but when you get it wrong, who, they let you know about it. So instead of taking the whole of suggesting how do you push that all through AI? We going to take that use case. I was speaking to a colleague in an online business. They're at 60% now all their customer service is on AI and they're getting better NPS, better stickiness, better revenue, every metric you can possibly think everything's everything is better. And we often think about ourselves. We want to create the experience that niche little pet store when you go in with your pet that the sawstore owner knows exactly what you need. They know. They know exactly. They're really tactile with what you do. We want to really try and recreate that experience. But you think about CS agents, we train them constantly, but I think we're up to now five hours a month, which is apparently quite a lot of training. But it all gets lost. You know, they go, they lose within an day. AI model, it's retained, it's retained and it improved. So you know huge applications I think there. Content I'm going to get 50% of our content out is AI generative. So of course product reviews, images, the whole gamut. That loads of applications there. I think personalization is work is actually the area I get really excited. I think we've talked about personalization as an industry for 20 years, like the mobile, you know, talk about it forever and ever. And when the iPhone came in it happened I think AI is going to be personalization at scale. That really gets me excited I think there are applications there that are just going to blow people away. And you know what's interesting is we talked about the issue of it not being human, I think AI is actually going to be the opposite. I think it's going to create really human interactions, really personal interactions. And at the moment, I'm still blasting a whole bunch of stuff out to people. I'm going to go do stuff maybe where I've got your image of your pet is actually, you know, laced through through your email with very specific personal recommendations of products and needs for that specific breed and age pet. So, you know, I'm really excited about that particular use case. You know, productivity loads. But I think actually with productivity we're finding start small like we got note taker and Google. Now every meeting has to have had note taken through Google. Great. You know, because you're allowing people to just start using the product or sort of seeing the value of AI. So I think that's important in analytics. We're starting to experiment with our own sort of analytics. So you know what, we're ourselves in margin last week or whatnot, etc., etc. so that you know, you've all sat there and tried to work your way through a hundred different tableau boards or in our case, Looker. I think that's very quickly going to be, you know, very specific LLM queries and you're going to get the data straight back. Just out of curiosity then, so you have a KPI to have 50% of images that would be gen AI or generated by AI, sorry. What timeline is that like I'm so curious. You've actually got some hard, hard metrics on this. June next year. June next year. All right. We're worried about most of our product reviews now go through AI. Most of our vet content. What happens is that vets write content and push it through generative AI. So we've thought about 10 or 15% now, but I think we'll get to I'm pretty confident we'll get to 50. I'm someone who absolutely has an unconditional bond with their dog. His name is Jett. He's a mini schnauzer. There's a couple of K that way. How you you know, you can't have my the name of my dog and the image of my dog without my consent and my permission. So AI and the large language models that we look at is only so good as the data that we're inputting. So what is your approach to zero party first party data? How are you capturing that and how are you leveraging Emarsys to do that? Fortunately, people love talking about their pets. So guilty. Yeah. So we haven't even started and 45% of customers are happy giving us that information. We've done a couple of tests like it's going to cost me maybe five bucks coupon to get to 100 really quickly with an image. So, I mean, honestly, I think it's so good. I think the mapping of that with the PIM data, the product information data will be our competitive moat. So if it'll cost me more, I'll do it because that to me everything else commoditized your web infrastructure, delivery networks, all that stuff it's ultimately commoditize. The marriage of those two pieces and if you get it right where I can make that almost like that local pet store recommendation is going to be a competitive moat and for me is that and that will be personalization of scale. And Maria, sort of similar question. I mean, we heard from Lisa Ronson earlier today. She used to be head of Coles. I walk into a Coles and buy Unilever product. I don't buy it from you. So how do you get data on your customer and how are you then feeding that into these models? Yeah, multiple touchpoints. I would say in the retail space that's a collaboration is something that we work quite a lot at scale. That's a collaboration with partners like Call 360, like Everyday Rewards and that sort of retail media space. That's something that we are developing. But yeah, of course we don't have such a reach that like sets us as a direct to consumer business. And we've been, but we've been testing with images on our retailer partners not as advance as you were saying but, and we've seen some rejection when you see a human that is not actually a human and it's generated through AI. So that's an interesting space that we are tapping into. So for example, we don't use hands that are not human hands. We just use AI to generate backgrounds, you do some specific things on products, but not leverage to everything and to everyone just yet because we're seeing a bit of backlash from our consumers. And we talked earlier about the touch point. Right. Where does AI end and the human touch remain? And my story from this is that we heard recently from a zero beer company. They were trying to understand how they promote zero beer in pubs. Right. Because it turns out when they spent time in the bars and asked bartenders, what are people asking for when they're not drinking when they're designated driver, it's just give me a water. It's fine. So their campaign was say no to water because that's actually who they're competing with, not other products and other drinks. But you don't really get that insight without spending time in the pubs. The data may not have surfaced. That right. So where's the break point, Jon, between the AI and trusting the data and getting the scale that we need to be a successful business versus that human impact? I mean, there's a great business quote that says, you know, we don't trust the data, look at the anecdotes. And so we spend a lot of time looking at CS interactions, for example, with customers. So very qualitative sort of data packets, if you like. Honestly, I mean, I've got to call out a little bit. I actually think that, as I said, I think we'll be able to do a better job of that in-store local pet store than than they will very quickly. So and I think that if you've got a pet, for example, what's your pet's name? Jed. Okay, Jed, let's talk about it. What what kind of breed is Jed? He's a black mini schnauzer. See how easy it is to get the information. What other zero party data do you want? I'm actually jealous, like, it's so easy. Unfortunately, I don't know enough about schnauzers. But anyway. But my point being is that let's say his joint issues are a problem, right? And glucose is really important. And your food doesn't have that. You're probably buying it very expensive premium diet. I suspect. So we will be able to say hey, actually Jed would benefit more from product D with glucose to me and so to me that's a really very personal insight that brings real value to you as a pet parent. So I actually think if you do this right this will be tactile and human in its delivery. Actually I'll rephrase that it needs to be tactile and human in its delivery. So you need to find that way that you're not. Gosh they just kind of made this stuff up and it's just all automated and it just needs to feel like it really is personal, really is about you and Jed. I think also the fail like it's unlikely to fail like if I go into a local pet store to give you an example and it's someone who I don't usually, you know shop from and he doesn't know that pet and the needs that's a breakpoint, right? So the data helps you kind of not fail that when I'm trying to buy a product that's not relevant for him. Yeah, absolutely. I mean, and you think about a store like say big box store is that they train their agents really thoroughly, but it breaks all the time, right? Like maybe you haven't had that training unit, so you're never going to get that kind of vast amounts of knowledge all stored into a singular interaction that you can through sort of an AI delivered solution. So, Maria, like where where are you seeing the break point? You know, where does the sort of creativity come in? Some of your products are fairly personal to people's needs. And so I'm sure there's also a high empathy radar on that, how are you guys thinking about this? Yes. The way that we think it's AI allow us to the volume like allow us to produce creative at scale allow us to deploy sort of like better digital marketing campaigns that may have like thousands of assets that we need to deploy. So AI provides that sort of like easiness and mass production adaptation of assets, creating copy by doing all of that. But the role of the marketers is, in my opinion, more than ever we really switch on with culture. So we are adopting an approach of like cultural first. So that's and that's the role of that we need to take us humans like direct AI to do things that like that can do and can do great. But taking the insights, understanding culture, like looking into trends, putting our brands in the front of society, like running our brands in the way that we want to run, that's probably the most important thing that my teams are doing at the moment, like immersing themselves in culture because all of the rest we have got are we have amazing tools that can facilitate. We have a suite of things that we can do but actually like building what our brands means. What are the roles in like our consumers life? That is not something that AI can do, at least in the short term. And I also think you rely on vendors capability, right? So how are we showing up to the party and producing a roadmap that allows you to be more bleeding edge in what you're doing? There's obviously multiple products in the market that are available but you want to double down on your existing partnerships as well to get there. I want to get some inspiration around impacts and results. So Maria, where are you seeing or what can you share that has really moved the needle in actually executing some of these campaigns using AI? Yeah, I think that we live across like a multiplicity of of categories. We have very good results on our shelf execution. For example, when you say AI but my favorite one, it's what we did with Dove recently we launched it's a 20 anniversary of the Dove selfie, the self-esteem campaign. And this year we went like really strong on what AI is actually doing on new generations and how they are sort of like interacting with that. It's really aligned with our sort of like Unilever view of the world of sustainable grow and leaving the planet better than we actually did. And we launched a toolkit for parents and for consumers. It's available in our website that educates the general public to be less biased, more like diverse in the way they prompt when they are prompting AI tools. And that was something when that was presented to me as a marketer, I was like, no one is going to understand this. Like, this is way too complex. Who's doing this? I mean, journey or interactivity, like who's actually doing this? But we decided to launch it anyways, and I was extremely surprised on the amount of parents calls consumers in general that downloaded that playbook to be less of a stereotype when generating A.I. images. So that's probably what I think that when we do well, when we think forward, when we think progressively, we see interaction and we see like good feedback on that. So that's something that's an initiative that really makes me proud from Unilever. That's excellent. And Jon, you're an Emarsys customer. You're using some of our features. So there high plug, is it a genuine use case? So we're using the omni channel personalization tool. We're still starting small and we've obviously got the name, but we also added now the image of the breed and we're testing out doubles the revenue percent. And again I'm looking we're scratching the surface on that. It could eventually be Jed, for example with very specific recommendations. So just through changing the imagery to map to the breed of the customer, that doubled the revenue percent. I think that scratches the surface of possibility. So, you know, excited to see. For me then the challenge is how do you take that? Because you still get to sort of create email. How do you create that and generate 66 breeds, you know, and so on. But I believe it will be possible. You know, I feel that I actually think in the personalization space, the notion of the campaign will go and you'll have segments on one side, objectives on the other, and then the machine sitting in the middle that will determine things like the images, you know, the coupons or the offer, you know, the channel like I think channels will go as a concept. I think you'll have it'll determine whether it should be, you know, Facebook, text, push email, whatever. So I think you know, that's scary for some people because my team spend most of the time building campaigns. But imagine then I go, I imagine this. Imagine if they thought about the customer segment and the outcomes and optimize to that. That's going to make them way more effective than sitting there building campaigns, it's like paying the Harbour Bridge, you know, start of the week, finish the end of the week, start again. And they don't really think about the customer. They don't really think about the results as such. So for me, this area gets me, it gets me super excited. I think that you know, the name of this global festival that Emarsys puts on is Power to the Marketer. And it's very intentional. We did an IDC report a year ago looking at customers that use the platform and it freed up 35% of marketers time that used the platform and also 53% quicker to get campaigns out, right? So that's pretty basic AI doing a lot of the heavy lifting that marketing teams do spend a lot of time on. And I think you just mentioned impacts, revenue. And I know Logan from Nighttime Bikes is somewhere here in the room. We just launched a case study with him last week. In the first 90 days of using the platform, they have seen an 8% increase in revenue. Right. So I think AI and the topic that we have just discussed today is important, but it also has to have an impact either on your people, your profit or, you know, really the customer experience, which then contributes to both of those things.
Leveraging AI for Growth: Future-Ready Marketing and Innovation
Watch on Demand
About This Webinar:
AI offers marketers incredible advantages—from inspiring creative design and crafting imaginative ad campaigns to delivering personalized recommendations and optimizing internal operations. But with so many possibilities, where should you start?
In this standout session from the Power to the Marketer Sydney, featuring speakers from Pet Circle, Unilever, and the Sydney Opera House, we’ll explore how some of the region’s most iconic brands are leveraging AI in practical, results-driven ways. Whether you’re looking for quick wins or long-term transformation, you’ll walk away with ideas you can put into action today.
What You Will Learn:
🔹 Enhancing Customer Service with AI: Discover how AI-driven solutions improve customer
interactions, streamline support, and enhance satisfaction.
🔹 Boosting Marketer Productivity: Learn how AI automates tasks, optimizes
campaigns, and enables teams to work smarter, not harder.
🔹 Democratizing Creativity: Understand how AI empowers marketers to
develop innovative content, personalize experiences, and scale creative execution efficiently.
Watch on Demand
Watch it now
Engage with the latest from the industry
Featured content
Real brands offering real customer engagement insights, including:
Personalize omnichannel engagement to build loyalty and
grow customer lifetime value
I have the pleasure of running the last panel today. I would like to ask a quick question. Actually, on behalf of the entire panel who has AI fatigue. Okay. So what we thought we'd do is end the day on AI. To help me do that, I would like to introduce Jon from Pet Circle and Maria from Unilever. Why don't we do a little bit of a round of introductions, maybe your name, where you're from and a little bit about, you know, where you've been in your career, if that's okay. Jon, we'll start with you. So just testing. Okay. I am Jon. I'm the Chief Growth and Marketing Officer at Pet Circle. I've been in digital marketing for longer than I care to admit, but for the last five years I've been working at Pet Circle, we're probably the largest e-com player that you've never heard of, I think we're the third largest by revenue. You know, we service almost a million customers annually. And what we're really trying to do is we're trying to create an experience where we can map the right product to every kind of pet species breed need. And so we're really trying to map the customer to the infinite amount of choice that exists out there. And of course, because we're an online, I can carry a lot more range than perhaps the brick and mortar competitors. And actually, the reason I ended up in digital is because 20 years ago I figured this thing called Google and Internet was going to be have a profound impact on marketing, I think AI is going to have even more of an effect. I think you can take AI, mobile and every other major societal event in the last 30 years, and A.I. is going to have that kind of impact. And maybe leave you with this intro. I think you got to have two mindsets when you're thinking about AI. First is you got to flip the narrative. So this is not the end of creativity. It's the start of infinite, more creative possibilities. Now, this is not the end of my teams. It's actually how you're going to actually enablement, empower them to do more. So I think you've got to really have a mindset that changes. This is going to enable a ton of stuff. And then, you know, I think the you know, the other so the other sort of key thing is you've got to start simple. It's really complex. So get out there And do I mean, I think I mean, one of the points made is you got to keep things simple because it's all about execution, you know, So and then I think actually the third I will add a third one is you got to educate yourself. There's a lot of information, a lot of misinformation. You know, I'm going back to school. I'm trying to learn as much about A.I. as I possibly can because it's complex and it's difficult to know where to start. So go out there and, you know, there's plenty of good, plenty of information and find out as much as you possibly can. Thanks, Jon. Maria, can we hear from you? Hard to come out like I'm after this big introduction, but I'll do my best. I'm Maria. I lead the digital marketing and data team for Unilever and I also lead the AI acceleration team for our ANZ business. I am not from Australia. I'm from Argentina. So I came to this country four years ago with my company with Unilever, and I been working at Unilever for the past five years here and in Latin America, and I do all of my career across marketing, digital marketing, data, a bit of sales and now dipping my toes into the retail media space and also the innovation across AI. I agree with Jon. I feel like we are living in this amazing era to increase our digital literacy, our AI literacy. So I'm really excited on everything that is happening and how we can use AI to do good things, to keep on doing good business and to keep on delivering on sustainable growth for our market. I think the first thing to acknowledge is that AI is not new. Silvana started today by acknowledging actually SAP Emarsys' has had AI in the platform since 2012. What I think a lot of us are really thinking about now is Generative AI, which is really the next iteration. And I'm going to steal a chain from Oroton's term, which I loved about being the author of the next chapter. Right. And Jon, as you just said, removing the ego and really thinking and being very intentional about what it means for your business and actually learning about the capability and the posture is super important. So my first real question is how is your business approaching AI top-down, bottom-up. How are you thinking about it and the approaches that you're putting in place? So I'll to acknowledge the first question I think it's got to be top-down. We tried initially trying to seed AI projects into the teams and through for many reasons. We just could not we could not get traction. I mean, there's the mindset. One, you know, if I adopt AI, I'm going to have a small team. There's a complexity piece. You know, there's so much information and then, you know, a day job. And so what we've done is we've made a top-down, we've got measures. We have a P1 system is one of my P1s. So I've actually got metrics and goals to get to drive AI into my areas. So that's one critical piece. I think the other piece is data. I mean, you've got to get your data right. It's been a lot. That's where you've got to spend a lot of time. In my business, we've got, we think about three tranches of data. And we have this debate. Who's our customer? Is it the person who buys the pet food and the pet stuff or is it the pet? Actually, we think it's the pet. So the most critical piece of the data. Three things I need to know. I really need to know two things plus one. It's nice. Need to know the breed, need to know how old it is. And ideally, I'd like to know the name so that we can start talking them personally. With those three things, I can make great recommendations. I can allow our customers to make great choices. The next piece is product information data. We go through this really laborious task now of re-tagging everything so that we can then eventually we can map these two things together. Right? So if I know you've got a Labrador, we know that I have joint issues. We also know that glucosamine is really good for joint issues. So I can then recommend food that has it. If I've done my tagging and my structures correctly so I can infer... I would actually, before you start getting into Generative AI and all this stuff, get your data right, think about how you're going to use it and get it right. And then the third, the least important piece of data is actually the person buying stuff, you know? But we need to know who they are. We need to know where they are, the demographics and all that stuff. But for me, the map of the pet with the product is really, really critical. Maria huge organization globally. I'm sure there's not a consumer that you haven't touched in terms of the product reach. How are you thinking about AI and the approach and the structure that you're thinking in a business? So in Unilever globally there's so much AI going on, it's probably we have a lot of fatigue for sure. There are over 500 initiatives running on AI across a multiplicity of the disciplines that we touch on. The approach is basically like two fold. We have sort of like a deep approach in innovation. So everything that AI can facilitate to provide a competitive advantage about the bandage. So projects that are like more IP protected, that are changing the way that we operate within our business, and then we have sort of like a more broad implementation that is what are the actual tools that everyone can use and we can sort of make available, train our teams, generate information towards that. But we have a very high ethical mindset on how we use AI and what is the data that we put in place. So of course putting our data together and making everything accessible, organized taxonomies and all of that, it's in place. So that's sort of like evergreen really, and basics. We go very like strict in how we sort of create and name our data sets. And then there is different approach around like specific implementations of tools. So yeah, like basic use of tools for productivity, for marketing, for data implementation. And then there is a massive work on innovation. And I think that in our cases, both like top-down and bottom-up, we have a page where we can sort of pitch for budgets and ideas and like new initiatives that we might want to use. And then there is a massive top-down from our global CEO on like embracing AI for productivity basis. So we are in kind of this like nice moment where everyone is sort of like getting excited and the people that is a bit afraid is getting trained to like sort of prevent that space. I think it's okay to be afraid. Yeah. And it's okay to normalize that. It's getting to another level of practicality than you said, starting to implement some some specific things. So in the lens of because AI can imagine a lot of different things. So worker productivity versus actual marketing. So in the context of today, what are some of the actual use cases and things that you're implementing now and the areas of focus? Yeah. So for me, everything that is happening on the digital commerce space is super important. Like how do you make sure that our omnichannel assets are created? Well, that's something that used to be very expensive and very complex for a company that has thousands of SKUs in market. Now we are using a lot of tools and AI to like gen AI on images to ensure that we have sort of like omnichannel coverage like perfectly done. We have a lot of productivity, of course, I feel like everyone is using Copilot or yeah, we are using Copilot like on our day to day basis to do like things that used to take me hours because I'm not from here and I'm hesitating my language a lot. Now copilot help me to do it better. And then we are working a lot on specifics of our campaign. So we have a very strong commitment to sort of expand the AI literacy into our consumers. So, for example, we have we just did a campaign that had a playbook to better promptings or real beauty prompting that is a tool that we are sharing in general, trying to educate everyone It's in there to avoid being bias and stereotype or preventing bad use. So yeah, lots lots going on from a marketing perspective. Lots going on. Jon same question. What are some of the sort of practical use cases or areas of focus that you're looking at? Yeah, the way that the top-down framework, we sort of split out the application into five areas that's content, customer service, productivity, personalization and analytics. So we've got owners for each of those areas. So to break down what we're doing. In our customer service as an example, keeping things simple. We we're going to just focus on the use case of where's my order when we get it wrong. It really, really pissed customers. Like now we've participated this unconditional bond between pet and pet parent. It's great, but when you get it wrong, who, they let you know about it. So instead of taking the whole of suggesting how do you push that all through AI? We going to take that use case. I was speaking to a colleague in an online business. They're at 60% now all their customer service is on AI and they're getting better NPS, better stickiness, better revenue, every metric you can possibly think everything's everything is better. And we often think about ourselves. We want to create the experience that niche little pet store when you go in with your pet that the sawstore owner knows exactly what you need. They know. They know exactly. They're really tactile with what you do. We want to really try and recreate that experience. But you think about CS agents, we train them constantly, but I think we're up to now five hours a month, which is apparently quite a lot of training. But it all gets lost. You know, they go, they lose within an day. AI model, it's retained, it's retained and it improved. So you know huge applications I think there. Content I'm going to get 50% of our content out is AI generative. So of course product reviews, images, the whole gamut. That loads of applications there. I think personalization is work is actually the area I get really excited. I think we've talked about personalization as an industry for 20 years, like the mobile, you know, talk about it forever and ever. And when the iPhone came in it happened I think AI is going to be personalization at scale. That really gets me excited I think there are applications there that are just going to blow people away. And you know what's interesting is we talked about the issue of it not being human, I think AI is actually going to be the opposite. I think it's going to create really human interactions, really personal interactions. And at the moment, I'm still blasting a whole bunch of stuff out to people. I'm going to go do stuff maybe where I've got your image of your pet is actually, you know, laced through through your email with very specific personal recommendations of products and needs for that specific breed and age pet. So, you know, I'm really excited about that particular use case. You know, productivity loads. But I think actually with productivity we're finding start small like we got note taker and Google. Now every meeting has to have had note taken through Google. Great. You know, because you're allowing people to just start using the product or sort of seeing the value of AI. So I think that's important in analytics. We're starting to experiment with our own sort of analytics. So you know what, we're ourselves in margin last week or whatnot, etc., etc. so that you know, you've all sat there and tried to work your way through a hundred different tableau boards or in our case, Looker. I think that's very quickly going to be, you know, very specific LLM queries and you're going to get the data straight back. Just out of curiosity then, so you have a KPI to have 50% of images that would be gen AI or generated by AI, sorry. What timeline is that like I'm so curious. You've actually got some hard, hard metrics on this. June next year. June next year. All right. We're worried about most of our product reviews now go through AI. Most of our vet content. What happens is that vets write content and push it through generative AI. So we've thought about 10 or 15% now, but I think we'll get to I'm pretty confident we'll get to 50. I'm someone who absolutely has an unconditional bond with their dog. His name is Jett. He's a mini schnauzer. There's a couple of K that way. How you you know, you can't have my the name of my dog and the image of my dog without my consent and my permission. So AI and the large language models that we look at is only so good as the data that we're inputting. So what is your approach to zero party first party data? How are you capturing that and how are you leveraging Emarsys to do that? Fortunately, people love talking about their pets. So guilty. Yeah. So we haven't even started and 45% of customers are happy giving us that information. We've done a couple of tests like it's going to cost me maybe five bucks coupon to get to 100 really quickly with an image. So, I mean, honestly, I think it's so good. I think the mapping of that with the PIM data, the product information data will be our competitive moat. So if it'll cost me more, I'll do it because that to me everything else commoditized your web infrastructure, delivery networks, all that stuff it's ultimately commoditize. The marriage of those two pieces and if you get it right where I can make that almost like that local pet store recommendation is going to be a competitive moat and for me is that and that will be personalization of scale. And Maria, sort of similar question. I mean, we heard from Lisa Ronson earlier today. She used to be head of Coles. I walk into a Coles and buy Unilever product. I don't buy it from you. So how do you get data on your customer and how are you then feeding that into these models? Yeah, multiple touchpoints. I would say in the retail space that's a collaboration is something that we work quite a lot at scale. That's a collaboration with partners like Call 360, like Everyday Rewards and that sort of retail media space. That's something that we are developing. But yeah, of course we don't have such a reach that like sets us as a direct to consumer business. And we've been, but we've been testing with images on our retailer partners not as advance as you were saying but, and we've seen some rejection when you see a human that is not actually a human and it's generated through AI. So that's an interesting space that we are tapping into. So for example, we don't use hands that are not human hands. We just use AI to generate backgrounds, you do some specific things on products, but not leverage to everything and to everyone just yet because we're seeing a bit of backlash from our consumers. And we talked earlier about the touch point. Right. Where does AI end and the human touch remain? And my story from this is that we heard recently from a zero beer company. They were trying to understand how they promote zero beer in pubs. Right. Because it turns out when they spent time in the bars and asked bartenders, what are people asking for when they're not drinking when they're designated driver, it's just give me a water. It's fine. So their campaign was say no to water because that's actually who they're competing with, not other products and other drinks. But you don't really get that insight without spending time in the pubs. The data may not have surfaced. That right. So where's the break point, Jon, between the AI and trusting the data and getting the scale that we need to be a successful business versus that human impact? I mean, there's a great business quote that says, you know, we don't trust the data, look at the anecdotes. And so we spend a lot of time looking at CS interactions, for example, with customers. So very qualitative sort of data packets, if you like. Honestly, I mean, I've got to call out a little bit. I actually think that, as I said, I think we'll be able to do a better job of that in-store local pet store than than they will very quickly. So and I think that if you've got a pet, for example, what's your pet's name? Jed. Okay, Jed, let's talk about it. What what kind of breed is Jed? He's a black mini schnauzer. See how easy it is to get the information. What other zero party data do you want? I'm actually jealous, like, it's so easy. Unfortunately, I don't know enough about schnauzers. But anyway. But my point being is that let's say his joint issues are a problem, right? And glucose is really important. And your food doesn't have that. You're probably buying it very expensive premium diet. I suspect. So we will be able to say hey, actually Jed would benefit more from product D with glucose to me and so to me that's a really very personal insight that brings real value to you as a pet parent. So I actually think if you do this right this will be tactile and human in its delivery. Actually I'll rephrase that it needs to be tactile and human in its delivery. So you need to find that way that you're not. Gosh they just kind of made this stuff up and it's just all automated and it just needs to feel like it really is personal, really is about you and Jed. I think also the fail like it's unlikely to fail like if I go into a local pet store to give you an example and it's someone who I don't usually, you know shop from and he doesn't know that pet and the needs that's a breakpoint, right? So the data helps you kind of not fail that when I'm trying to buy a product that's not relevant for him. Yeah, absolutely. I mean, and you think about a store like say big box store is that they train their agents really thoroughly, but it breaks all the time, right? Like maybe you haven't had that training unit, so you're never going to get that kind of vast amounts of knowledge all stored into a singular interaction that you can through sort of an AI delivered solution. So, Maria, like where where are you seeing the break point? You know, where does the sort of creativity come in? Some of your products are fairly personal to people's needs. And so I'm sure there's also a high empathy radar on that, how are you guys thinking about this? Yes. The way that we think it's AI allow us to the volume like allow us to produce creative at scale allow us to deploy sort of like better digital marketing campaigns that may have like thousands of assets that we need to deploy. So AI provides that sort of like easiness and mass production adaptation of assets, creating copy by doing all of that. But the role of the marketers is, in my opinion, more than ever we really switch on with culture. So we are adopting an approach of like cultural first. So that's and that's the role of that we need to take us humans like direct AI to do things that like that can do and can do great. But taking the insights, understanding culture, like looking into trends, putting our brands in the front of society, like running our brands in the way that we want to run, that's probably the most important thing that my teams are doing at the moment, like immersing themselves in culture because all of the rest we have got are we have amazing tools that can facilitate. We have a suite of things that we can do but actually like building what our brands means. What are the roles in like our consumers life? That is not something that AI can do, at least in the short term. And I also think you rely on vendors capability, right? So how are we showing up to the party and producing a roadmap that allows you to be more bleeding edge in what you're doing? There's obviously multiple products in the market that are available but you want to double down on your existing partnerships as well to get there. I want to get some inspiration around impacts and results. So Maria, where are you seeing or what can you share that has really moved the needle in actually executing some of these campaigns using AI? Yeah, I think that we live across like a multiplicity of of categories. We have very good results on our shelf execution. For example, when you say AI but my favorite one, it's what we did with Dove recently we launched it's a 20 anniversary of the Dove selfie, the self-esteem campaign. And this year we went like really strong on what AI is actually doing on new generations and how they are sort of like interacting with that. It's really aligned with our sort of like Unilever view of the world of sustainable grow and leaving the planet better than we actually did. And we launched a toolkit for parents and for consumers. It's available in our website that educates the general public to be less biased, more like diverse in the way they prompt when they are prompting AI tools. And that was something when that was presented to me as a marketer, I was like, no one is going to understand this. Like, this is way too complex. Who's doing this? I mean, journey or interactivity, like who's actually doing this? But we decided to launch it anyways, and I was extremely surprised on the amount of parents calls consumers in general that downloaded that playbook to be less of a stereotype when generating A.I. images. So that's probably what I think that when we do well, when we think forward, when we think progressively, we see interaction and we see like good feedback on that. So that's something that's an initiative that really makes me proud from Unilever. That's excellent. And Jon, you're an Emarsys customer. You're using some of our features. So there high plug, is it a genuine use case? So we're using the omni channel personalization tool. We're still starting small and we've obviously got the name, but we also added now the image of the breed and we're testing out doubles the revenue percent. And again I'm looking we're scratching the surface on that. It could eventually be Jed, for example with very specific recommendations. So just through changing the imagery to map to the breed of the customer, that doubled the revenue percent. I think that scratches the surface of possibility. So, you know, excited to see. For me then the challenge is how do you take that? Because you still get to sort of create email. How do you create that and generate 66 breeds, you know, and so on. But I believe it will be possible. You know, I feel that I actually think in the personalization space, the notion of the campaign will go and you'll have segments on one side, objectives on the other, and then the machine sitting in the middle that will determine things like the images, you know, the coupons or the offer, you know, the channel like I think channels will go as a concept. I think you'll have it'll determine whether it should be, you know, Facebook, text, push email, whatever. So I think you know, that's scary for some people because my team spend most of the time building campaigns. But imagine then I go, I imagine this. Imagine if they thought about the customer segment and the outcomes and optimize to that. That's going to make them way more effective than sitting there building campaigns, it's like paying the Harbour Bridge, you know, start of the week, finish the end of the week, start again. And they don't really think about the customer. They don't really think about the results as such. So for me, this area gets me, it gets me super excited. I think that you know, the name of this global festival that Emarsys puts on is Power to the Marketer. And it's very intentional. We did an IDC report a year ago looking at customers that use the platform and it freed up 35% of marketers time that used the platform and also 53% quicker to get campaigns out, right? So that's pretty basic AI doing a lot of the heavy lifting that marketing teams do spend a lot of time on. And I think you just mentioned impacts, revenue. And I know Logan from Nighttime Bikes is somewhere here in the room. We just launched a case study with him last week. In the first 90 days of using the platform, they have seen an 8% increase in revenue. Right. So I think AI and the topic that we have just discussed today is important, but it also has to have an impact either on your people, your profit or, you know, really the customer experience, which then contributes to both of those things.